The overall aim of this proposal is to understand and model the processes that underlie the formation of conscious, reportable memories of otherwise fleeting visual representations. When we are looking for something, we can pick out that target surprisingly quickly when our eyes light on it. This is shown by presenting items one after the other at rates of 10 or even 20 a second using rapid serial visual presentation (RSVP). Such experiments with sequences of letters, words, or pictures have shown that the first target is easy to spot, but then it is strikingly difficult to detect a second target for the next half second (500 ms), an effect called an attentional blink. A complex sequence of processes occurs in that 500 ms, starting with detection of the first target, followed by a brief period of enhanced attention called transient attention which rapidly draws attentional resources to the location of the target. After this initial burst of attention, the blink occurs, making it very difficult to spot a new target anywhere in the visual field for 200-500 ms afterwards. Surprisingly, this blink is entirely eliminated by presenting an unbroken stream of targets. Thus a string of targets keeps attention engaged, but if there is a gap in the targets, attention is switched off. This effect suggests that there is a struggle between top down and bottom up control for attention. To understand these dynamics, a computational model of the temporal control of attention is explored in collaboration with B. Wyble. The eSTST model produces predictions that are tested with a series of 18 experiments aimed at exploring the competitive regulation of attention in several domains, including single streams of characters, spatial arrays, repetitions, words, pictures, and sentence reading. Computational modeling of the sort described here makes potentially informative links with disorders of attention because the model relates behavioral phenomena to neurally plausible structures. As a result, the model's account of attentional control can be related to anatomically specified structures in neuroimaging studies, as well as to temporally specified brainwave signals in EEG studies. The attentional processes and their limits that we propose to study have broad importance for many practical situations in modern life in which the visual environment changes rapidly, including city driving, air traffic control, and response to medical emergencies. [unreadable] [unreadable] [unreadable]